Building AI Agents for Finance: a new book from the Finance Loop network

Finance Loop network member Fayssal El Mofatiche has co-written a book, and it lands right in the middle of what the Finance Loop network works on. Building AI Agents for Finance, written with Hanane Dupouy and published by Packt in August 2026, is a hands-on guide to building agents that hold up in real financial systems.

AI agents are changing how financial systems read information, reach decisions and automate work that used to need a person in the loop. Plenty of material explains agentic AI at the concept level. Far less of it shows how to build agents that run reliably once money and regulation are involved. That is the gap this book sets out to close.

For Finance Loop, that makes it an easy one to recommend.

The book starts from the ground up: what an AI agent is, how it differs from a non-agentic system, and when an agentic architecture is the right call instead of a plain workflow. From there it moves into Python labs built on Claude and OpenAI models, working through design patterns, memory, retrieval-augmented generation, framework choice and reasoning methods such as ReAct, Reflection, Self-Consistency, Language Agent Tree Search and multi-agent collaboration.